Drop'ed is a sophisticated automation suite built for the Concordia Deckathon 2026. It is designed to navigate the complex "Concordia Deckathon" student portal, solving over 12 unique anti-bot challenges and CAPTCHAs to automate the student dropout process.

The project consists of three main components:

1. Automation Core: A high-performance pure API script (pureApiDropout.js).
2. Desktop Application: A user-friendly Electron app for easy execution (frontend/).
3. Landing Page: A marketing website for the tool (website/).

For the fastest exection, run the file pureApiDropout.js directly. This script uses a direct API approach, using only the browser for showing the completed drop out message.

1. Install Dependencies

   npm install

2. Configure Environment Create a .env file in the root directory (copy from env_example.env):

   GEMINI_API_KEY=your_gemini_api_key_here
   CHROME_PATH=C:\Program Files\Google\Chrome\Application\chrome.exe

   Note: A valid Gemini API key is required for solving the Face CAPTCHA modules.

3. Run the Script

   node pureApiDropout.js

We provide a standalone desktop application for a simpler user experience.

Location: /frontend

cd frontend
npm install
npm run dev

To create a standalone .exe file that you can share:

cd frontend
npm run pack

This generates frontend/dist/Dropped 1.0.0.exe.

The project includes a promotional landing page.

Location: /website

cd website
npm install
npm run dev

This tool automates the entire student lifecycle in the challenge environment:

1. Account Creation: Generates random credentials and registers via API.
2. Authentication:
   • Login: Bypasses logic puzzles and CAPTCHAs.
   • MFA: Intercepts OTP codes sent to the user's email.
3. Class Management: Identifies enrolled classes and drops them via API.
4. Financials: Checks balance and performs credit card payments if there are outstanding fees.
5. Dropout: Finalizes the dropout process after solving the "Pretty Faces" verification.

The script successfully navigates 12 distinct verification modules:

We employ a multi-modal approach to solving the varied CAPTCHAs:

• Pixel Analysis (Logos): We use sharp to analyze pixel brightness histograms to detect white/black center pixels in logo grids.
• Generative AI (Crowds): We utilize Google's Gemini Flash 1.5 to visually identify humans in the "Select all images with people" challenges.
• Brute Force (Final): For the final challenge, we use a parallelized batching strategy to brute-force the 512 possible combinations (2^9) in seconds.

• Node.js: Core runtime.
• Electron: Desktop GUI framework.
• Puppeteer Real Browser: For mimicking human fingerprints and evading detection.
• Google Gemini AI: For vision-based CAPTCHA solving.
• Sharp: High-performance image processing.
• React + Vite: Frontend framework for the website and electron app.

Made with love for ConUHacks 2026